This is the first year of a three-year continuing Research Initiation Award. The research aims to develop a new kind of adaptive machine for the orientation and presentation of small parts in a manufacturing assembly system. Such machines have the potential for a large impact in manufacturing, especially in small- batch manufacturing in the context of agile manufacturing systems. The research explores the fundamental science of automation required to build such a machine, including development of appropriate methods for reasoning under uncertainty, machine learning methods for design of feeder track reorienting devices, and methods of automated planning for design of new feeder tracks. The ideas will be tested on a prototype software-programmable parts feeder for laminar parts, based on vibratory bowl feeder designs.
